The Thermal Contact Resistance Dataset and the Artificial Intelligence-Driven Prediction of Thermal Contact Resistance in Multi-material Systems

分类： 动力与电气工程 >> 工程热物理学 分类： 计算机科学 >> 计算机应用技术 提交时间： 2025-04-11

Man Zhou Pei Yao Guo Zhuo Yan He Ping Zhang Tao Lin

摘要： In the current era of artificial intelligence, the advancement of high-performance computing based on electronic devices is hindered by thermal contact resistance. To accurately predict this resistance, we established a comprehensive database derived from extensive experimental work documented in previous studies. By employing machine learning algorithms, we developed a prediction model for thermal contact resistance that utilizes this dataset. This model can predict the thermal contact resistance among all learned materials, demonstrating a significant degree of general applicability. Our model shows strong performance on the test set (with a coefficient of determination of 0.982) , reflecting a high level of predictive accuracy. Additionally, the interpretability analyses conducted on the machine learning model are consistent with established theories of thermal contact resistance, further confirming the model’s accuracy. We anticipate that this database will support the development of thermal contact resistance prediction models and that our model will enhance the precision of thermal contact resistance predictions.

同行评议状态:待评议

 点击量 819  下载量 97  评论 0